The key point about this tech? If you're using your own art assets, it's usable for commercial movie-making, with no license fee.

Any of the learned HN types know how the Source engine stacks up graphics algorithms-wise with other engines like Unreal 3 and Crysis? Specifically, how are their lighting standards, and do they support Global Illumination / Indirect Lighting stuff, and/or Ambient Occlusion?

Update - apparently they do support SSAO, but on a per-material basis. Not ideal, but hey. And they do Indirect Lighting, but baked, not dynamic.

We use Houdini's digital asset technology to seamlessly develop two version of each asset: one for the gaming engine, and one for the film-quality renders. We use the gaming engine to do all of the camera work, set dressing, initial lighting setups, etc. We can easily create any shot, just like Pixar does with their animated films. We use this ability to iterate on the story in the context of the fully edited film as we constantly increase production value in each scene.

Then, when a scene is solid, we move on to film-quality lighting and rendering. Our facial capture is done at 120fps with up to four RED Epic cameras, each with special stereo hardware we've developed (so up to eight 5.5 MP images are captured simultaneously 120 times a second). Along with the mo-cap data for actors, this allows us to finish the film entirely within the computer.

We're currently looking into Crytek's engine, and I guess we'll look into Source Filmmaker now too. It's a great time to be making films with the computer.

Sadly, the VC industry is completely dry in this area. I wonder what's up with that?

I look forward to seeing some high quality content in the near future, and will hopefully create some myself.

http://www.youtube.com/watch?v=TXnXOn7WRCs

The graphics seem a lot more realistic, but the Source Filmmaker might be easier to use.